Lesson 6 of 7
A familiar shape from the mainstream
A mainstream economist argued that policy rates in the 2000s sat below a simple rule and housing responded. The claim, the counterfactual, the limits.
By Javier Sanz
Founder of ValueScreener. Built Ninety Nine, a securities services provider, and sold it to a tier-one global bank; ran operations and internationalization at Alpaca, the brokerage API.
Who writes thisEditorial processLinkedIn (opens in a new tab)X (opens in a new tab)
In one minute
- A policy rule is a benchmark, not a law. It says what a rate would be if the central bank responded to inflation and output as it had before.
- The argument is that the US policy rate ran well below that benchmark in the early 2000s and that housing responded strongly.
- The evidence is a simulation from an estimated equation, which is a model of one market and not a measurement of what happened.
Readers of this track will recognise the shape of this argument: a rate held below a benchmark, a boom concentrated in long-lived assets, a bust. What makes it worth its own lesson is that it was made by a mainstream economist using published data, and that it can be examined rather than asserted.
What the argument has and what it lacks
What is on the table
- A published benchmark rate path, computed from a rule.
- The actual federal funds rate, which sat below it for several years.
- An estimated relationship between the funds rate and housing starts.
- A simulation showing a smaller boom under the benchmark path.
What is not
- Any claim that the rule is the correct policy.
- Any measurement of what would actually have happened.
- Any account of the lending standards and securitisation of the period.
- Any test that separates this explanation from others.
A counterfactual from one estimated equation is evidence about that equation as much as about the world.
The claim
John Taylor's 2007 paper, prepared for a central bank symposium, argues that the federal funds rate in the years after 2001 was held below what a simple policy rule of the kind he had earlier proposed would have implied, and that this was the largest such deviation since the 1970s.
He then asks what difference it made. Rather than assert an answer he estimates a housing starts equation on the federal funds rate using quarterly data from 1959 to 2007, reports an interest rate semi-elasticity of about minus 8.3, and simulates the path of housing starts from late 2000 under a counterfactual funds rate that follows the rule.
The simulated path tracks the data while the actual policy is close to the rule, and separates afterwards: on the counterfactual path the increase in housing starts is much smaller. His conclusion is that a higher rate path would have avoided much of the housing boom according to this model.
What that does and does not establish
It establishes that one estimated relationship, fitted over five decades, implies a materially smaller boom under a different rate path. That is a real result and it is checkable, because the inputs are published series and the equation is stated.
It does not establish causation in the episode. A single-equation simulation holds everything else fixed, including the lending standards, the securitisation of mortgages and the global demand for dollar assets that other accounts put at the centre. Taylor's own paper describes the approach as more straightforward than the international model with rational expectations that the question ideally needs.
And it is not a rule about what policy should be. A policy rule in this literature is a description of how a central bank has responded, used as a benchmark against which a deviation can be measured. Treating the benchmark as the right answer is a separate step that this lesson does not take and the track does not permit.
Why it belongs in this course
Because it is the nearest thing to the Austrian claim that a rate held below a sustainable level distorts the pattern of investment, made in the mainstream's own language, with a published series and a stated equation.
The similarities are real and so are the differences. There is no structure of production in the argument, only one long-lived asset class. There is no claim that the boom had to end in a particular way. And the benchmark is an econometric regularity rather than a natural rate derived from time preference, which is exactly the concept Sraffa's objection questioned two courses ago.
Check yourself
What is a policy rule used as, in this argument?
- A legal requirement on the central bank
- A benchmark describing how the central bank has responded before, against which a deviation can be measured
- A forecast of future rates
Show the answer
Answer: B.
The deviation is the observation. Treating the benchmark as the correct policy is a further step and is not part of the evidence.How did Taylor estimate the effect on housing?
- By surveying builders
- By estimating a housing starts equation on the federal funds rate over 1959 to 2007 and simulating a counterfactual rate path
- By comparing US housing with European housing
Show the answer
Answer: B.
He reports a semi-elasticity of about minus 8.3 and simulates from late 2000, with the counterfactual path following the rule.What does the simulation not hold constant?
- Nothing, it holds everything constant
- It does hold other factors constant, which is why lending standards and securitisation are outside the result
- The federal funds rate
Show the answer
Answer: B.
A single-equation counterfactual varies the rate and holds the rest fixed, so any explanation resting on those other factors is neither tested nor excluded.Sources
- Taylor, 'Housing and Monetary Policy', NBER Working Paper 13682, December 2007, prepared for a Federal Reserve Bank of Kansas City symposium: the size of the deviation of the federal funds rate from a policy rule in the early 2000s, the housing starts equation estimated on quarterly data from 1959 to 2007, the semi-elasticity of about minus 8.3, the counterfactual simulation from late 2000, and the author's own statement that a fuller answer needs an international model with rational expectations.
What this leaves out
- This is a heterodox tradition. The mainstream account of the same phenomenon differs, and nothing here is a forecast or a view on any security.
- The paper's earlier statement of the rule itself was not obtained for this lesson, so the rule is described only as it is used here.
- Nothing in this lesson is a policy position on interest rates, past or present, and nothing in it is a claim about any current rate path.